Where To Find The Latest AI Research? Top 7 Sources to Stay Updated

Whether you are an AI engineer, a student, a researcher, a developer, or simply an AI enthusiast, you might have felt the fear of missing out on the latest advancements in AI, with so much happening every week. Staying informed of cutting-edge developments can give you a competitive edge and keep you updated with the rapidly evolving field. The good news is that, unlike many other research domains, almost all recent AI research papers are available online for free.

This article will explore the different journals, conferences, platforms, communities, and approaches to find the latest AI research papers. Everyone has their preferences. Skim through and explore these resources and build your preference for staying updated in AI research, or just bookmark this page for later.

7 Sources: Summary Table

Source	Description
Preprint Archives	Platforms like ArXiv offer immediate access to cutting-edge research before peer review. Tools like Arxiv Sanity help manage the large volume of papers.
Search Engines for Research	Tools like Google Scholar, Semantic Scholar, and Consensus provide search and summaries for academic papers.
Using AI to Find Papers	AI-powered tools like SciSpace, Consensus, and ChatGPT plugins help discover and summarize research efficiently.
Top Journals	Peer-reviewed journals like Neural Networks, IEEE Xplore, and Journal of Artificial Intelligence Research provide high-quality AI research.
Company Research Blogs	Blogs from companies like OpenAI, DeepMind, and Google AI share applied AI findings.
Top Conferences/Workshops	Events like NeurIPS, ICML, and CVPR showcase the latest research and breakthroughs in AI.
Platforms and Communities	Platforms like PaperswithCode, ResearchGate, and HackerNews offer tools, networking, and updates.

Source #1 Preprint Archives

Preprint archives are platforms where researchers publish their papers before they are peer-reviewed, published in different journals, or presented at a conference, which usually takes months. Platforms such as ArXiv.org have gained popularity in the AI community because they provide immediate access to the latest research, allowing the AI community to engage with new ideas at an early stage. Access to preprint has increased visibility and citations of open-access papers, enabling open community feedback and collaboration. You can check the recent pages in ArXiv for AI, ML, computation and language, Neural and Evolutionary Computing, Computer Vision, and Pattern Recognition to stay updated with the topic of your interest.

However, an overwhelming number of papers have been published on ArXiv. Andrej Karpathy created Arxiv Sanity Preserver to help researchers manage this by providing an effortless way to track, sort, and search for documents based on similarity. A lite version of this platform is also available at arxiv-sanity-lite.com. You can create an account, browse or search for specific topics, and explore similar papers to get started. When you find documents of interest and save them to your library, Arxiv Sanity provides personalized recommendations and updates. Here’s a video by Karpathy on how to use it:

SSRN(Social Science Research Network) is another preprint server that focuses more on social science domains, including humanities, life sciences, and health sciences. This site can be handy if you are looking for cross-domain research papers.

Source #2 Search Engines for Research Papers

Many journals and conferences publish enormous amounts of research papers, and various search engines and platforms aggregate research papers and make it easier for you to discover relevant work in AI.

Google Scholar is the most widely used and accessible search engine for academic papers in all domains. You can also use power search techniques such as Boolean operators and keyword optimization to find specific documents. For example,

Furthermore, you can create email alerts in Google search, such that Google Scholar will send you a regular update of new research by an author on a particular topic you selected.

Alternatively, platforms, such as Semantic Scholar, use semantic search to find relevant publications based on the topics. Furthermore, it can provide super-short summaries of the objectives and results of the paper with its TLDR features and recommend articles based on your document browsing history.

Another platform, Consensus, more focused on general domains, has broad search capabilities and a Consensus Meter that aggregates findings from multiple papers to show the overall Consensus on a particular topic, helping you gauge the general agreement in the research community. It also has an integrated copilot feature that offers AI-powered assistance for drafting content, creating lists, and answering questions based on research papers.

Source #3 Using AI to find Papers on AI

Why not ask an AI to give you papers on AI? This way, you could quickly narrow down your research with appropriate context in your prompt. Plugins available in ChatGPT, such as SciSpace, Consensus, and ScholarAI, help you find papers based on your prompt. Tools like consensus and scispace are already utilizing ChatGPT in their platform to get answers, however we can use them directly through ChatGPT as well. To get started, click on Explore GPTs in your chatGPT account. Search for the plugin, and let's use scholarAI. Type your prompt and start the chat. You might have to sign in / sign up for the platform and allow ChatGPT to talk with the plugin.

You can customize it with better prompting techniques; however, the plugin is optimized to give you titles, summaries, and DOI links to the article and a downloadable PDF link (if available).

If your preferred LLM has access to the web, you can use it too, but the results might not be as good as using a database-linked LLM. In this regard, perplexity.ai can help you combine multiple papers toward your research goal, allowing you to narrow down your research through academic papers.

Source #4 Top Journals

Academic journals provide peer-reviewed research to ensure a high standard of quality. Many such journals, such as IEEE Xplore, ACM (Association for Computing Machinery) Digital Library, may require subscriptions. Many of these journals are well-regarded and ranked with different methodologies. Here are some of the top journals:

• Neural Networks covers research on neural networks and related learning systems and has been in publication since 1988.
• Foundations and Trends in Machine Learning have the highest impact factor (IF), covering a wide range of topics in ML.
• International Journal of Computer Vision has the highest IF in the field of computer vision.
• Artificial Intelligence publishes papers on broad aspects of AI.
• IEEE Transactions on Pattern Analysis and Machine Intelligence publishes research in artificial intelligence, particularly in AI, machine learning, and computer vision.
• Journal of Artificial Intelligence Research is an open-access journal publishing high-quality research across all domains of artificial intelligence.

These are just a few of thousands of journals.

The question is, how can you find top journals in your domain of interest? You can use a journal finder such as Elsevier Journal Finder or search on reputable journal databases. A few of these journal databases are listed below:

• Web of Science by Clarivate, which also acquired the ProQuest Database, provides access to multiple databases across various disciplines.
• IEEE Xplore by the Institute of Electrical and Electronics Engineers (IEEE) offers computer science, electrical engineering, and electronic engineering materials.
• Scopus by Elsevier is well regarded in most academic domains.
• JSTOR is a digital library offering access to academic journals and books.

Many of these journals are paid and require subscription; only a few allow open-access. However, you can get access to this research through alternative approaches.

Source #5 Company Research Blogs

Unlike in some other domains, major technology companies are pivotal in publishing groundbreaking work in AI. These companies often have dedicated research blogs sharing their latest findings, white papers, making them an excellent source for applied AI research. Here's a list of a few blogs from the major players:

• OpenAI: focuses on researching LLMs and aims towards artificial general intelligence.
• DeepMind focuses on solving challenging problems in various domains (games, genomics, proteomics, robotics, and so on) with AI.
• Google AI focuses on machine intelligence, responsible AI, scientific discovery, generative AI, and foundation models among other research interests.
• Meta AI: Focuses on AI research related to the Multiverse, which can be vision, language, audio processing, 3D, generative models, etc.
• NVIDIA focuses on hardware, different AI applications, and AI training aspects. Following these blogs can keep you updated on the particular technology space.

Source #6 Top Conferences/Workshops

Annual conferences heavily influence AI research, where the latest breakthroughs are often presented before being formally published in journals. Many of these conferences publish their proceedings online, making them accessible to a broad audience. Here’s the list:

• NeurIPS (Conference on Neural Information Processing Systems) covers a broad spectrum of AI research, including deep learning, neural networks, and computational neuroscience.
• ICML (International Conference on Machine Learning) is known for presenting state-of-the-art research in ML
• CVPR (Conference on Computer Vision and Pattern Recognition) and the European Conference on Computer Vision (ECCV) feature pioneering work in image recognition, object detection, and computer vision technologies.
• AAI Conference on Artificial Intelligence focuses on both theoretical and applied aspects of AI from a broad perspective
• The International Conference on Learning Representations (ICLR) highlights the latest research in representation learning, including deep learning and unsupervised learning.
• International Conference on Robotics and Artificial Intelligence (ICRAI) for robotics and artificial intelligence research.
• The IEEE International Conference on Artificial Intelligence (ICAI) focuses on AI applications in key verticals such as healthcare, life science, engineering, business intelligence, safety, etc.
• ACL (Association for Computational Linguistics) is one of the popular conference in the field of natural language processing (NLP) covering different aspects of computational linguistics

Source #7 Platforms and Communities

In addition to different journals, conferences, blogs, search engines, there are certain platforms and communities that can help us discover relevant AI research papers and help us stay updated.

Paperswithcode provides top, new, popular research and models and codes for various ML problems. Getting access to the code of the top model for a particular benchmark can speed up your research even further.

ResearchGate is a social networking site for scientists and researchers to share papers, ask and answer questions, and find collaborators. Such a social networking platform can be beneficial when journals are behind a paywall, and you can directly request the author for access to the papers and even other resources if available for sharing.

HackerNews is a social news website focusing on computer science, technology, and entrepreneurship. Y Combinator, a well-known startup accelerator, runs it. It has a minimalistic design to prioritize substance over form, where users can submit stories related to these topics, which get upvoted by the community. It can be a great news source about the latest AI research and tools.

Conclusion

AI is evolving rapidly; the techniques and algorithms we use today can be outpaced by another in a matter of days, making current techniques, algorithms, and models obsolete. Hence, staying updated with the latest AI research is mandatory if you are working in any cutting-edge domain.

You can ensure you never miss out on groundbreaking advancements by leveraging various sources like preprint archives, top journals, conferences, and specialized search engines. Furthermore, To make your research process even more efficient, consider using Large Language Models (LLMs) like ChatGPT with plugins such as SciSpace, Consensus, and ScholarAI. These tools can help you find relevant papers quickly and provide context-rich summaries, making it easier to stay informed and ahead in the fast-paced world of AI research. Start exploring today and transform how you stay updated with AI advancements!

You can cite this work as follows:

@article{ResourcesLatestResearch2024Shikhrakar,
  Title = {Where to Find the Latest AI Research Papers?},
  Author = {Rojesh Shikhrakar},
  Year = {2024},
  url={https://learnprompting.org/blog/resources_latest_research_papers}
}